Re: CPU Shares
by Marcos Sungaila
Hi Nick,
Setting CPU Shares only changes how the instance works. It does not affect the current workload unless the instance is under pressure.
For example, if you are running an RDBMS, you may see a performance improvement when normalizing the database or generating a monthly report. Even that, the improvement may be minimum.
Do not expect a sensible performance improvement for normal operation.
Performance tuning is not magic. You will often perform many configurations to see a 1%, maybe 2% gain.
Regards,
Marcos
From: Nick --- <jsmith1299(a)live.com>
Sent: Wednesday, February 1, 2023 9:59 AM
To: users(a)ovirt.org
Subject: [External] : [ovirt-users] CPU Shares
Hello,
I would like some help understanding CPU shares. We have several VMs which are CPU pinned due to licensing. We would like to favor one VM since it is our application VM over the other VMs which it is pinned with. Would I set the CPU share on the one that needs to priority to high or low?
In the documentation I see the following:
Specify CPU Shares. Possible options are Low, Medium, High, Custom, and Disabled. Virtual machines set to High receive twice as many shares as Medium, and virtual machines set to Medium receive twice as many shares as virtual machines set to Low. Disabled instructs VDSM to use an older algorithm for determining share dispensation; usually the number of shares dispensed under these conditions is 1020.
I currently have it set to high and I believe it is doing the opposite of what I want it to do as the CPU steal percentage appears to be 10-15% over the other VMs which show no CPU steal and are set to low. I'd like to confirm if I have this setup correctly or not.
Thanks,
Nick
1 year, 10 months
Re: Self-hosted engine 4.5.0 deployment fails
by John
I would run the deploy again, wait until the engine is up and then from
the server you are deploying the engine on
# virsh list
obtain the virtual machine number from the above command, in the example
below we assume the number is 1
# virsh console 1
login as root
wait for all packages to finish updating:
tail -f /var/log/dnf.rpm.log
then:
# dnf downgrade postgresql-jdbc
once that's done the deployment of the engine should get past the
[ INFO ] TASK [ovirt.ovirt.engine_setup : Check if Engine health page is
up]
step.
On 03/05/2022 17:03, Mohamed Roushdy wrote:
>
> Hello,
>
> I’m deploying on Ovirt nodes v4.5.0.1, but it fails to deploy the
> self-hosted engine with the following errors in the web installer:
>
> [ INFO ] skipping: [localhost]
> [ INFO ] TASK [ovirt.ovirt.engine_setup : Check if Engine health page
> is up]
> [ ERROR ] fatal: [localhost -> 192.168.222.197]: FAILED! =>
> {"attempts": 30, "changed": false, "connection": "close",
> "content_encoding": "identity", "content_length": "86",
> "content_type": "text/html; charset=UTF-8", "date": "Tue, 03 May 2022
> 15:57:20 GMT", "elapsed": 0, "msg": "Status code was 500 and not
> [200]: HTTP Error 500: Internal Server Error", "redirected": false,
> "server": "Apache/2.4.37 (centos) OpenSSL/1.1.1k
> mod_auth_gssapi/1.6.1", "status": 500, "url":
> http://localhost/ovirt-engine/services/health}
> [ INFO ] TASK [ovirt.ovirt.engine_setup : Clean temporary files]
> [ INFO ] changed: [localhost -> 192.168.222.197]
> [ INFO ] TASK [ovirt.ovirt.hosted_engine_setup : Sync on engine machine]
> [ INFO ] changed: [localhost -> 192.168.222.197]
> [ INFO ] TASK [ovirt.ovirt.hosted_engine_setup : Set destination
> directory path]
> [ INFO ] ok: [localhost -> localhost]
> [ INFO ] TASK [ovirt.ovirt.hosted_engine_setup : Create destination
> directory]
> [ INFO ] changed: [localhost -> localhost]
> [ INFO ] TASK [ovirt.ovirt.hosted_engine_setup : include_tasks]
> [ INFO ] ok: [localhost]
> [ INFO ] TASK [ovirt.ovirt.hosted_engine_setup : Find the local
> appliance image]
> [ INFO ] ok: [localhost -> localhost]
> [ INFO ] TASK [ovirt.ovirt.hosted_engine_setup : Set local_vm_disk_path]
> [ INFO ] ok: [localhost -> localhost]
> [ INFO ] TASK [ovirt.ovirt.hosted_engine_setup : Give the vm time to
> flush dirty buffers]
> [ INFO ] ok: [localhost -> localhost]
> [ INFO ] TASK [ovirt.ovirt.hosted_engine_setup : Copy engine logs]
> [ INFO ] changed: [localhost]
> [ INFO ] TASK [ovirt.ovirt.hosted_engine_setup : Change ownership of
> copied engine logs]
> [ INFO ] changed: [localhost -> localhost]
> [ INFO ] TASK [ovirt.ovirt.hosted_engine_setup : Notify the user about
> a failure]
> [ ERROR ] fatal: [localhost]: FAILED! => {"changed": false, "msg":
> "There was a failure deploying the engine on the local engine VM. The
> system may not be provisioned according to the playbook results:
> please check the logs for the issue, fix accordingly or re-deploy from
> scratch.\n"}
>
> Any idea if this is a bug or something? I tried to deploy for eight
> times with different configuration
>
> Thank y9ou,
>
>
> _______________________________________________
> Users mailing list --users(a)ovirt.org
> To unsubscribe send an email tousers-leave(a)ovirt.org
> Privacy Statement:https://www.ovirt.org/privacy-policy.html
> oVirt Code of Conduct:https://www.ovirt.org/community/about/community-guidelines/
> List Archives:https://lists.ovirt.org/archives/list/users@ovirt.org/message/XL...
1 year, 10 months
CPU Shares
by Nick ---
Hello,
I would like some help understanding CPU shares. We have several VMs which are CPU pinned due to licensing. We would like to favor one VM since it is our application VM over the other VMs which it is pinned with. Would I set the CPU share on the one that needs to priority to high or low?
In the documentation I see the following:
Specify CPU Shares. Possible options are Low, Medium, High, Custom, and Disabled. Virtual machines set to High receive twice as many shares as Medium, and virtual machines set to Medium receive twice as many shares as virtual machines set to Low. Disabled instructs VDSM to use an older algorithm for determining share dispensation; usually the number of shares dispensed under these conditions is 1020.
I currently have it set to high and I believe it is doing the opposite of what I want it to do as the CPU steal percentage appears to be 10-15% over the other VMs which show no CPU steal and are set to low. I'd like to confirm if I have this setup correctly or not.
Thanks,
Nick
1 year, 10 months
Cluster level upgrade (from 4.3 to 4.6) and Windows VM
by duparchy＠esrf.fr
Hi,
When trying to upgrade the Cluster level to 4.6, a popup panel says "Due to the emulated machine change, some Windows virtual machines might lose static IP configuration"
It is not clear if Windows VMs might be disconnected from the network right away when clicking OK or only when restarted...
Bets regards
1 year, 10 months
